Over the past 15 years, quantitative trait locus (QTL) mapping has identified hundreds of chromosomal regions containing genes affecting atherosclerosis or other disease-related phenotypes in mice, yet the underlying genes and pathways have remained largely elusive. During the present grant cycle, we helped develop a systems-based approach, which we term integrative genetics, that is proving useful in not only identifying the genes underlying QTL but also in elaborating the complex genetic and environmental interactions in traits such as atherosclerosis. The fundamental concept is to use common genetic variation, as it exists among inbred strains of mice, to help organize whole genome expression array data into biologically relevant networks that link to both DMA variation and clinical trait variation. During the present ' cycle, we applied this approach to certain metabolic and cardiovascular traits, using expression array data from tissues such as liver, muscle, and adipose. The results were highly encouraging, as we developed networks that predicted novel genes for obesity and vascular calcification. The goal of the present proposal is to adopt this integrative genetics .approach to study the interactions of,vascular cells, macrophages, and lipids as they relate to atherogenesis. Our work thus far has utilized linkage analyses of data from crosses between inbred strains of mice. We now propose to complement this with association analyses of a mouse diversity panel, consisting of about 100 inbred strains that have been largely sequenced. Such a panel has important advantages for mapping resolution and for integrating physiologic/pathologic measures requiring analysis of multiple mice. Recent data, generated since the first submission, provide strong proof of concept evidence for the approach. In order to examine atherosclerosis and related traits among the panel, we propose the use of a dominant hyperlipidemia model that will be bred to each of the 100 strains to create genetically diverse heterozygous mice for phenotypic analyses. Our proposal is organized into three interactive Aims that integrate linkage analyses (Aim 1), association analyses (Aim 2), and mathematical modeling (Aim 3).